Using Science Superpowers

This is the bootstrap that makes the whole skill system work. It loads first in any conversation and enforces a simple rule: if there's even a 1% chance a skill applies to what you're doing, invoke it before responding. No clarifying questions first, no quick peeks at the data, no rationalizing that this one thing is too simple. It includes a flowchart showing when to invoke skills and a table of red flag thoughts that mean you're about to skip the discipline. The instruction priority is clear: your explicit instructions override everything, then skills, then default behavior. Worth reading the red flags table, it's basically a field guide to how analyses go wrong.

Install to Claude Code

npx -y skills add k-dense-ai/science-superpowers --skill using-science-superpowers --agent claude-code